So we’re finally up to the capital letters! Luckily, capital letters are much easier to design than their lowercase counterparts. You can get functioning glyphs for a third of the uppercase alphabet with just a day’s work. This is because uppercase letters are much more square and geometric than their lowercase counterparts. There are fewer curves to worry about and the curves that do exist are much easier to handle. Many uppercase letters also look very similar and recycle parts (at least optically), so a few parent letters can give rise to very large swaths of the alphabet.
The most fundamental unit of the capital letters is the vertical stem. That means that the first letter you should design should be the ‘I’.
It makes sense to design capital letters where they would most commonly be used—at the beginning of a word, so that’s typically my approach to designing capitals.
The ‘I’ can be constructed by taking a lowercase ‘i’ and duplicating its bilateral serif. The ‘I’, unlike the ‘i’, is perfectly symmetrical, so make sure the right side of the serif isn’t still shorter than the left side.
It’s a myth, perpetuated by those elementary school
handwriting rules, that capital letters extend to the same height as the ascenders of lowercase letters like ‘d’. In reality, capitals are almost always a bit shorter, usually tucked underneath the bottoms of the head serifs of the ascenders (some typefaces like News Gothic make them the same height though).
Capital stems are also a bit wider than lowercase stems, and their serifs a bit longer and thicker. It is simply because capitals are so much larger than lowercase letters that if they kept the same stroke width, they would look thinner than the lowercase letters (their black area increases linearly with letter size while their whitespace increases
quadratically). The capital–lowercase stem width ratio is governed by two conflicting interests. On one side, the eye wants to see the same color density between the two cases (meaning the capitals would just be scaled-up versions of lowercase letters), on the other, the eye wants to see consistency between the two stroke widths. For some designers, the ratio can be as high as 6:5, though personally I find that distracting and so I go for a much lower ratio.
The capital ‘O’ is essentially a scaled up version of the lowercase ‘o’, corrected for weight. Just like with the ‘I’, the absolute width of the hairline and bowl should be slightly wider than in the lowercase ‘o’. Just like the lowercase letters, round uppercase letters have overshoots (the capital ‘I’ is provided for reference).
The ‘O’ is also much closer to a perfect circle than its lowercase variant is. That means it’s slightly wider than a scaled up ‘o’, and also a little wider.
The ‘V’ is similar, though I recommend just extending the arms up to the capital line (determined by the height of the ‘I’) and emboldening slightly rather than performing a scalar transformation on it. The diagonal should be the same width (or very slightly less) as the stem of the ‘I’. Remember to make the serifs wider and thicker.
The capital ‘Z’ can be constructed the same way too. Try to keep the top arm shorter than the bottom arm, just like in the lowercase ‘z’. (Yes, “Zebra”—there was no western state, province, or country I could think of that starts with ‘Z’. The ‘V’ is there for reference since the two letters are quite similar.)
And the same set of corrections should be applied. The diagonal on the ‘Z’ should be the same width as the diagonal on the ‘V’, and the same goes for the arms. The serifs need to be larger, as usual, but they are also less angled than in the lowercase ‘z’.
And that concludes our first four capital letters!
